àáâ ãäåæç - mediaTUM

Transcription

àáâ ãäåæç - mediaTUM
TUM
I N S T I T U T F Ü R I N F O R M A T I K
State of the Art in Mobile Social Networking on
the Web
Georg Groh, Philip Daubmeier
TUM-I1014
August 10
T E C H N I S C H E U N I V E R S I T Ä T M Ü N C H E N
TUM-INFO-08-I1014-0/1.-FI
Alle Rechte vorbehalten
Nachdruck auch auszugsweise verboten
c 2010
Druck:
Institut für Informatik der
Technischen Universität München
State of the Art
in Mobile Social Networking
on the Web
Georg Groh and Philip Daubmeier
Institut für Informatik
Technische Universität München
November 2009
1
Abstract
This contribution presents a study on currently existing mobile social network (MSN) platforms.
We briefly discuss the basics of mobile social networking and defines our view on related terms
and concepts. We then discuss and develop a classification pattern for MSN platforms and services. Several such systems are arranged and grouped, according to this categorization scheme.
Finally, an empirical survey is presented, pointing out key dynamics, similarities and differences
between the presented systems.
2
3
CONTENTS
Contents
1 Introduction
4
1.1
Mobile Social Networking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4
1.2
Goals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5
1.3
Outline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6
2 Concept disambiguation
2.1
Categorization design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3 Classification
6
8
10
3.1
Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
10
3.2
MSN: Mobile Versions of ’Classic’ Social Networking Platforms . . . . . . . . . .
10
3.3
Microblogs
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
11
3.3.1
Classic Microblogs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
11
3.3.2
Location Based Microblogs . . . . . . . . . . . . . . . . . . . . . . . . . .
11
MSN: Distinctly Mobile Social Networking Platforms . . . . . . . . . . . . . . . .
12
3.4.1
Messaging and Chatting . . . . . . . . . . . . . . . . . . . . . . . . . . . .
12
3.4.2
Interest Aware . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
12
3.4.3
Multimedia Sharing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
12
Peer-to-peer mobile social networks . . . . . . . . . . . . . . . . . . . . . . . . . .
12
3.5.1
SMS based . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
13
3.5.2
SMS based and location aware . . . . . . . . . . . . . . . . . . . . . . . .
13
Near field mobile social networks . . . . . . . . . . . . . . . . . . . . . . . . . . .
13
3.6.1
Wi-Fi based networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
14
3.6.2
Social situations vizualising systems . . . . . . . . . . . . . . . . . . . . .
14
Multiple social network access and instant messaging . . . . . . . . . . . . . . . .
15
3.7.1
Location aware versions . . . . . . . . . . . . . . . . . . . . . . . . . . . .
16
Location based mobile social networks . . . . . . . . . . . . . . . . . . . . . . . .
16
3.8.1
Friend finding services . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
16
3.8.2
Locating . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
18
3.8.3
Tracking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
18
3.8.4
Meeting new people . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
19
3.8.5
Interest awareness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
19
3.8.6
Friend finding, location sensitive content . . . . . . . . . . . . . . . . . . .
20
3.8.7
City Guide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
21
3.8.8
Nightlife . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
21
3.4
3.5
3.6
3.7
3.8
1
4
INTRODUCTION
3.8.9
3.9
Trips . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
22
3.8.10 Alerting when friends are nearby . . . . . . . . . . . . . . . . . . . . . . .
22
3.8.11 Awarding systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
22
3.8.12 Storytelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
23
3.8.13 Location and distance aware bulletins . . . . . . . . . . . . . . . . . . . .
23
Mobile social networks with microblogging . . . . . . . . . . . . . . . . . . . . . .
24
3.9.1
Interest and location aware . . . . . . . . . . . . . . . . . . . . . . . . . .
24
3.9.2
Location based geojournals . . . . . . . . . . . . . . . . . . . . . . . . . .
24
3.9.3
Multimedia sharing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
24
4 Empirical Analysis
25
4.1
Social Network Crawler . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
25
4.2
Network dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
26
4.2.1
Development of mobile social network start-ups . . . . . . . . . . . . . . .
29
4.3
Geographic dependencies in utilisation . . . . . . . . . . . . . . . . . . . . . . . .
29
4.4
Spread and usage of different mobile platforms . . . . . . . . . . . . . . . . . . .
29
4.4.1
29
Usage of mobile platforms . . . . . . . . . . . . . . . . . . . . . . . . . . .
5 Conclusion
32
5.1
Future approaches in MSN
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
32
5.2
Completeness of this contribution . . . . . . . . . . . . . . . . . . . . . . . . . . .
35
1
Introduction
Mobile Social Networking (MSN) is an important field in mobile and social computing. The
rich possibilities for acquiring individual and social context in the mobile domain of use and
the interesting new classes of services that allow for usefully augmenting social life with added
value services represent a natural further development of the paradigm of desktop based Social
Networking as realized in platforms as Facebook. It is not the goal to review and cover all the
approaches in scientific literature (such as context aware computing, social signal processing,
reality mining etc.) that may contribute to or are relavant for mobile social networking. The
goal of this contribution is to analyze the current state of the art of mobile social networking
services and platforms existing on the (mobile) Web as of September 2009.
1.1
Mobile Social Networking
Social Networking (SN) platforms such as Facebook [1] or MySpace [2] are among the most
popular applications or platforms on the Web as of early 2010 (see e.g. [3]). These platforms
evolved from community platforms (with their bundle of communication-, information- and
awareness-services [4]) by adding explicit models of social networks and corresponding service
1
INTRODUCTION
5
bundles. Many of the most popular SN-platforms also offer UIs specially designed for mobile access. But merely interacting with Facebook’s Website from an I-Phone should not be considered
as mobile social networking. Social Networking, in general, can be defined as a collaborative or
social computing paradigm, that
• makes use of an explicit social network model
• (selectively) opens parts of the participating user’s information spaces to other users
from the social network
• offers useful communication-, information-, and awareness-services that can operate
on these parts
The concept of Mobile Social Networking can be defined as an extension or broadening of
this paradigm by
• detecting and modeling the user’s individual and social context
• using these context models and selectively opening them (as part of the individual information space) to other users from the network to enhance the service bundle and increasing
its usefulness especially in mobile interaction scenarios
The social context e.g. encompasses dyadic social relations on small time and space-scales
or models for social situations while the individual context encompasses current location etc..
Detecting and modeling social relationships on small space- and time-scales is the realm of the
emerging field Social Signal Processing [5, 6, 7]. Earlier approaches [8, 9, 10] limited the
definition of mobile social networking merely to SN platforms with mobile access. While mobile
access patterns to ”conventional´´ SN-services may certainly differ in some aspects from desktop
interaction with the same SN-services, e.g. with respect to the context-sensitivity (e.g. spatiotemporal relevance / validity) of the communicated contents, the bundle of services remains
basically identical and does usually not use individual and social contexts in an innovative way.
Another innovative field that is related, with respect to some aspects, to mobile social networking
is decentralized social networking [11, 12, 13, 14]. Decentralized social networking gives
up central servers hosting the social networking platforms and relies on Peer to Peer (P2P)
paradigms allowing each user to be socially “represented” by a peer (or even an agent). Besides
the general advantages of P2P (such as scalability, reliability etc.), distributing social networking
solves the cross-platform problem of not being able to easily span modeled social structures,
context model instances and other entities across platform boundaries in a natural way. In
decentralized social networks each user autonomously manages her profile (including her ties in
the SN, access control etc.), which gives back more control to the user. Besides the advantages,
in decentralized SN some problems may become harder than in centralized SN, e.g. the problem
of group-formation and demaraction [15]
1.2
Goals
While the definition of mobile social networking of the previous section is well founded, it
is mainly oriented towards the future of mobile social networking. The goal of this paper is
a survey of existing mobile social networking approaches, platforms or systems on the Web
2
CONCEPT DISAMBIGUATION
6
today. With “existing” we mean such approaches, platforms or systems (commercial or not) in
at least roughly beta-stage that that are open to the public and that have a non-trivial user-base,
meaning that it is not limited to the developers only. This may include research prototypes of
a certain maturity.
We aim at including all approaches, platforms or systems that
• implement key aspects of social networking and
• offer a mobile UI
in order to allow for a broad classification of these systems in view of establishing, justifying
and developing an agreeable definition of mobile social networking. In other words, the goal of
this paper is to thoroughly characterize the current state of MSN in order to be able to identify
and characterize existing problems and possibilities for future developments and research in this
field on the basis of this characterization. It provides a basis on which the desirable concept
definition for MSN given above builds on.
1.3
Outline
In section 2 present a classification system for mobile social networking services and platforms
(according to their most prominent services) is presented and several options for such a classification schema are discussed. In section 3 the available mobile social network platforms and
services are classified according to the classification system and for each a short discussion is
given, describing key features and properties. Section 4 presents empirical analysis of the dynamics of key properties of mobile social network platforms such as the growth rates of the social
network, distribution with respect to operating systems, geographical distribution etc. Section
5 concludes this contribution by discussing future trends and research challenges.
2
Concept disambiguation
For today’s mobile social networking, three main concepts play an important role:
• Weblogs, or simply Blogs (see e.g. [16] for an early discussion).
• Location based services (see e.g. [17] for a general overview on context-aware systems
including LBS).
• Social networking (see 1.1).
Figure 1 shows flavors and combinations of these concepts with a mobile and non-mobile interaction paradigm. Euclidean distance is intended to resemble the conceptual similarity between
the depicted concepts as far as possible.
At the intersection of blogs and social networks, Microblogs were placed. A prominent example
is Twitter [18]. These systems allow users to publish small messages about what they are
currently doing, or dealing with. Each user has its own channel that can be subscribed by others
(‘following’). While many Blogs are organized on separate Web-pages in a decentralized way,
2
7
CONCEPT DISAMBIGUATION
SN Social networks
M
Classic
SN
MSN
Microblog
Mobile
microblog
Mobile
Blog
Loc. based
microblog
Mobile
loc. based
microblog
Mobile
loc. based
Blog
Loc. based
SN
loc. based
MSN
Mobile
loc. based
service
Mobile services
Mobile websites
Mobile phone applications
Blog
Weblogs
LBS Location based services
Figure 1: Three main concepts, their combinations and their mobile versions.
implementing 1:n anonymous communication, microblogging is featured mostly on centralized
platforms such as Twitter. Through the directed ’follows’ relations, which party de-anonymizes
the 1:n communication, a loose social network is formed, which justifies the placement in the
diagram.
In the figure, MSN (Mobile (access to) Social Networking) represents mobile access versions of
classic (desktop-oriented) social networking (SN). These versions include adapted mobile UIs
(e.g. adapted Web-pages or -applications or native mobile applications) that go slightly beyond
merely accessing the given UI (e.g. Web-pages) with a micro-browser from a UI perspective but
that do not offer substantially new services that make use of mobile context.
Four concepts that can be justified to comprise the current state of mobile social networking are
illustrated in figure 2, all lying in the cross section of ‘mobile’ and ‘social network’ (MSN, Mobile
microblog, Mobile location based microblog and location-based MSN). While other classes of
figure 1 like e.g. Blogs may also exhibit aspects of social networking and can, in principle, also be
accessed and interacted with via a mobile device (e.g. via a micro-browser), we limited ourselves
to systems, approaches and platforms that fall into one of the depicted four classes.
Social networking platforms usually offer a range of services, UIs and extension mechanisms
which are also shown in figure 2:
• A web interface, that lets users access and manage their accounts and profiles and access
the platforms SN-services by using a web browser.
• A mobile web interface (MSN) (see above).
2
8
CONCEPT DISAMBIGUATION
Webinterface
provides
Mobile
Webinterface
APIs
M
Classic
SN
MSN
Microblog
Mobile
microblog
Mobile
Blog
Loc. based
microblog
Mobile
loc. based
microblog
Mobile
loc. based
Blog
Loc. based
SN
loc. based
MSN
Mobile
loc. based
service
Mobile services
Mobile websites
Mobile phone applications
provides
Authentication
API
Social network
API
Blog
Weblogs
used to connect to other
provides extensibility through
accessed by
SN Social networks
accessed by
Applications (Programs)
Mobile
Applications
LBS Location based services
Applications (Plugins)
Figure 2: Mobile social networking systems
• An API (Application programming interface), allowing third party applications to access
to the service. Furthermore it may allow other social networks to partly reuse existing login
information and connect together different accounts of a single user. It is also possible to
automatically synchronize particular data between the connected networks, such as status
updates, which are then changed in both systems synchronously. Such an API typically
consists of the two fundamental parts:
– An Authentication API, which verifies login data and grants access.
– A social network API, which allows for reading and changing profile data (including
social network data).
2.1
Categorization design
During the investigation of existing mobile social networks, many platforms and service-bundles
were found. Figure 1 already provided a coarse introductory grouping / classification of approaches relevant to mobile social networking. In order to more thoroughly structure and differentiate between the various approaches, besides the classification schema of figure 1, several
other / augmentable classification schemes are imaginable:
1. Differentiate by the underlying social networking model:
• Are edges of the network directed, weighted or typified or not?
• Are friends appended automatically, or only manually? Is a confirmation mandatory?
2
CONCEPT DISAMBIGUATION
9
2. Classify according to privacy concepts implemented.
3. Group by statistical measures such as:
• Popularity
• Dynamics of the system or vitality/activity of users
• Attendance or just unique visitor numbers
• User count
4. Partition technologically by used implementation techniques, standards, APIs, programming languages, platform toolkits, etc...
5. Group by provided key features and services (information-, communication or awarenessservices)
6. Group by the degree of provision of semantic meta-data (amount of usage of Social Semantic Web [19][20])
7. Group by the fact whether it is centralized or decentralized.
8. Group by the degree of interoperability with other platforms.
Schema 1 is inappropriate for our overview, because almost all existing social network platforms
or mobile social network approaches use undirected unweighted and untypified relationships,
with manual adding mechanisms for friends via two-way handshakes. The few exceptions that
have directed friend relationship models are mostly small and new systems, or rather simple implementations. Weighted relationships are also not implemented yet in any of the systems found
that have a non-trivial user-base. Typifying edges of the users network is also not implemented
in most of the currently existing services. Nearly all exceptions of this rule only provide a simple
friend grouping / friend-list mechanism such as in Facebook, with certain possible implications
for privacy (managing information access levels). Few other systems (Groovr, see 3.8.6 and
Whrrl, see 3.8.12) offer the possibility of assigning friends to one of two categories: the ‘circle
of trust’ and ‘friends’. Privacy settings can then be set separately for each of these types.
Schema 2 (structuring according to privacy) is also not very expressive. Observations showed
that there are mainly two types of social network services: Ones that provide privacy features
like hiding elements of the profile to non-friends, and ones that do not provide any privacy
options.
Schema 3 takes several empirical values into account. We will discuss these statistical data in a
separate section. However, most of these statistical measures are not suitable for an expressive
classification for our purposes since the goal of our classification, as it was stated in 1.1, is to
thoroughly characterize the current state of MSN in order to be able to identify and characterize
existing problems and possibilities for future developments and research in this field.
Schema 4 is mainly only interesting in view of the elements of schema 5 to 9. The mere technological implementation details are not interesting in view of the goals of this survey.
Schema 6 and 7 are indeed interesting but at the time of this survey, almost no platforms with
non-trivial user-base use. Social Semantic web or operates in a decentralized way.
Thus, we classify the approaches according to the schema in 1 and use augmented characterizations 5 and 8 where appropriate.
3
10
CLASSIFICATION
3
3.1
Classification
Sources
To create this overview of mobile social networks, besides our own research we used Web-sources
listing social networks like [21, 22], location-based services [23] and mobile social applications [24,
25]. Moreover, several app-stores were investigated w.r.t. mobile social applications, such as the
iTunes application store, Android market and the Ovi store by Nokia. Many systems had to be
filtered out, which were labeled as social networks, but turned out to be just chatting, messaging,
recommending or multimedia sharing systems without any social networking abilities.
3.2
MSN: Mobile Versions of ’Classic’ Social Networking Platforms
In this section we very briefly review established social networking platforms that also have a
mobile version (either versions of the Web-pages optimized for mobile access or a stand-alone
mobile Application). Since these platforms are well known and have been broadly discussed, we
refrain from discussing them in a very detailed way.
Facebook [1] is the most popular social networking site on the web [3]. Facebook’s ‘connect’ mechanism is widely used by other social networks, and a multitude of applications are available, which use the Facebook’s API to extend
the social network.
Friendster [26] is one of the first social network platforms, gone online in
2003, and is very popular in Australia and Asia. This social network platform
offers, aside from common functions, just few rather exceptional features. Friendster allows
artists, musicians, organizations and other celebrities to create fan pages, where Friendster users
can add themselves as ‘fans’. Since 2008, when the mobile web interface was launched, users
can get SMS text alerts on friend requests, new messages etc [27]. Friendster offers, just like
Facebook, a rich API, supports OpenSocial [28] and holds a myriad of applications (plugins),
that extend functionality of the user interface).
MySpace [2] is one of the largest social network platforms at the time of
assembly of this paper and was launched 2003. Since 2006 it provides a mobile web interface. There are lots of mobile applications for virtually every
platform either from MySpace itself or made by third party developers. MySpace is known for
its customizability of the personal profile page of a user, by letting users decide what colors,
background pictures and style-sheets ought to be used. Furthermore, the platform is famous for
having a lot of musicians and bands registered, allowing artists to upload their songs and get in
contact with their fans.
Hi5 [29] is one of the most popular classic social networking platforms, according to their unique visitors (see section 4). The service was launched 2003,
and has over 70 million users. [30]. The mobile webpage was launched in 2008.
3
11
CLASSIFICATION
Hyves [31] is another classical social network platform, launched in 2004. It
is foremost popular in the Netherlands, with about two thirds of the about 9
million total users being Dutch.
Orkut [32] is a social networking platform acquired by Google, launched in
2004, which is very popular in Brazil (more than the half of all users [33]), and
for this reason managed and operated by Google Brazil since 2008. Orkut has
a mobile web interface since 2008.
Bebo [34] is a classical social network platform since 2005. It was acquired by AOL in 2008 [35]. It provides a mobile web interface [36] since
2007.
3.3
3.3.1
Microblogs
Classic Microblogs
Twitter [18] is the most popular microblogging service today, and the place
13 of the most visited sites in the world (according to the Alexa traffic rank
[37]). It is also often referred as social network, but in this paper, is to be
classified as pure microblogging system (see section 2). Twitter messages can be sent via SMS
from the mobile phone, or by logging in to the web interface on a PC or a mobile browser.
It provides a rich API which can be used by a multitude of third-party applications that are
available on all platforms of mobile devices.
3.3.2
Location Based Microblogs
Location based microblogs make use of the geo-location, a user currently resides in.
Zannel [38] is a location based mobile microblogging service with support
of photo and video attachments. The location from where the message was
sent can be set by a #@location keyword in a SMS, or is set by the mobile
application by using the phones positioning abilities. Currently, the only supported platform for
a native application is iPhone OS, but Zannel can be accessed by a mobile web interface too,
supporting all mobile browsers.
Shizzow [39] is a location based microbloging system allowing users to ‘shout’
from a specific place. Users can not only see shouts of people they have subscribed to, but also geographically nearby shouts. Shouts can either be posted
and read by the mobile web interface, or by the Android mobile application. Additionally messages can be sent by SMS, either containing the current location or the service uses the last
known position of the user.
3
12
CLASSIFICATION
3.4
MSN: Distinctly Mobile Social Networking Platforms
In this section we review social networking platforms that are mainly targeted for mobile use.
We classify them according to their main purposes and services.
3.4.1
Messaging and Chatting
Mokomobi [40] is a mobile social networking platform, offering not only
messaging services but also live chats, group chat rooms and photo and video
sharing. Mobile applications exist for J2ME enabled platforms and iPhone OS.
Furthermore, a mobile Web interface is available, as the URLs top level domain
(‘.mobi’) already suggests.
Qeep [41] is a German mobile social network platform, featuring a J2ME
mobile application, with which users can send text messages, play games, share
photos and send so called ‘sound attacks’ - sounds that are played on the target
phone after having received it.
3.4.2
Interest Aware
Aka-aki [42] is a German social network, online since 2007. It features a
J2ME based and an iPhone mobile application. Users can add ‘stickers’ to
their profiles to communicate their interest to guests. Additionally, a so called
‘stick-o-meter’ shows in a bar visualization how many of the users stickers match to the own
ones and therefore tells the user how many interests are shared.
Seamee [43] is an Austrian mobile social network, offering a J2ME mobile
application to access the service. Seamee is divided into the four subnetworks
‘job’, ‘business’, ‘people’ and ‘love’. Users can decide in which of these subnetworks they want to participate, or take two or more of them at a time.
3.4.3
Multimedia Sharing
Next2friends [44] is a mobile social network that allows to share photos and
live video streams and offers mobile applications for J2ME, Blackberry OS and
iPhone OS platforms.
3.5
Peer-to-peer mobile social networks
Peer-to-peer social networking platforms let peers (e.g. mobile devices) themselves autonomously
handle communication between them. In contrast to that, conventional platforms store users’
messages on their servers waiting to be retrieved by the respective recipients. For this reason,
privacy issues with respect to the social network operator or potential attackers that infiltrated
the system may occur.
3
CLASSIFICATION
13
Peer-to-peer social networks do not exhibit this weak point, because messages are stored on the
recipients’ devices ins distributed way. Furthermore it does not demand much storage space
and saves bandwidth of network operators’ servers, if multimedia files such as photos and videos
are sent directly to other users, rather than to be stored on central servers. The following
two network systems work with a SMS based communication channel. Systems using IP based
communication, however, potentially will appear in the near future as there are already such
systems for desktop computers, such as Noserub [13] for example.
3.5.1
SMS based
Cellmate [45] is a peer-to-peer mobile social network that is based on a
J2ME application that uses the mobile phones short message service (SMS)
for communication between two Cellmate applications. Therefore it does not
need a central server, not even for brokering between the users, and no registration. Once the
application is installed on the mobile device, it asks for mobile phone numbers of friends that can
be organized in groups afterwards. This way, a completely decentralized social network is built
up, with users only knowing of their direct friends. Conversations are displayed chronologically
like in instant messaging programs. For planning an event, users can send checklists to all
members of a group, or create polls which results are sent automatically back to the creator of
the survey. Notifications for events can be created, that alarm all relating people. Because of
the SMS based transmission channel, this application can be quite expensive for the user.
3.5.2
SMS based and location aware
Are you here? [46] is a location aware peer-to-peer mobile social network
based on the platform ‘Wanted Smiling’, released by Clicmobile [47]. By sending ‘ON’ via SMS to the service, the user is set as visible in the system and his
location is shared with other people around him. The service sends back a list
of people that are near and sorts them according to the particular relationship:
friend, friend of friend, and so on up to 6 degrees of acquaintanceship. The communication
between the users then goes on by SMS. Therefore the network just functions as an agent to
localize near people, while the communication works on a peer-to-peer basis by using the short
message service (SMS).
3.6
Near field mobile social networks
This category lists social networks, which avail themselves of wireless technologies, working only
in small distances up to several meters. Such technologies are for instance Bluetooth [48] and
Wi-Fi. Some of the following systems are peer-to-peer based. Nevertheless the peer-to-peer
services were divided into the two categories ‘Peer-to-peer mobilesocial networks’ and ‘Near
field mobile social networks’, because of the completely different usage patterns that go along
with the systems. Peer-to-peer systems in the previous section aim to provide a messaging
platform, regardless of location and time. ‘Are you here?’, which is a location based network,
allows its users to still communicate, even if they are not near to each other, and even if one
participant has switched off his device. The SMS message, in this case, will arrive when the
3
CLASSIFICATION
14
phone is switched on again. The systems shown below, however, are strongly dependant on time
and the users’ place.
3.6.1
Wi-Fi based networks
iFob [49] is a location aware mobile peer-to-peer social networking application. It just works within reach of WiFi hotspots, like in coffee shops. Users do
not have to register on any central server, they just have to enter a ‘tagline’,
which is used as a conversation opener, for example ‘I like coffee’. This tagline
is broadcasted in the Wi-Fi network and received by any other iFob users. ‘iFob
will become your beacon, sending out a local signal which says you might want company. Or
set your copy of iFob so it just listens, showing you the other iFob users while letting you chose
whom you might want to chat with’ [49].
3.6.2
Social situations vizualising systems
Jabberwocky is a location aware mobile social software. It is a project of members of the
Urban Atmospheres group within Intel Research Berkeley [50]. The mobile application, executable in J2ME environments, uses Bluetooth technology to log every nearby mobile device,
which has Bluetooth switched on. Other people do not have to have the application installed,
but just have to set their phones to be visible by other Bluetooth devices. The application
now vizualises all persons in the proximity, likely beeing strangers, by drawing a red square on
the upper edge of the display. As time passes, and the previously detected device gets out of
reach, the square slowly moves to the bottom of the display and finally fades out after a while.
Using this visualization, shown in figure 3, the user
can see at a glance when he was near strangers, and
how many they were. Additionally he sees where he
met them, because he knows where he was at a specific
time. If a stranger is detected, that the user encountered before, it will appear as a green square on the display, indicating a so called ‘familiar stranger’. The visualization takes into account the number of strangers
nearby, the number of encounters with each stranger,
the time spent near a stranger and the elapsed time between encountering a stranger again [52]. Jabberwocky
thus builds up a social network, having weighted and
typified edges, and saves it in a decentralized manner
Figure 3: Jabberwocky mobile applicaonly locally on the phone. In [51] the Jabberwocky
tion (Source: [51])
developers describe two exemplary scenarios where the application can be useful:
• A person, new to a city, does not feel at home. Jabberwocky reassures him that people are
nearby, he already encountered, without even knowing about it, and so grants confidence.
• A person feels like the large city seems more like a small town, after years of living and
working there, and wants to escape the daily routine in his spare time. As he walks along,
he can see on his mobile device how many familiar strangers are around him. Having found
an unknown place, he feels much less crowded and explores new places in the city.
3
15
CLASSIFICATION
Figure 4: Encounter Bubbles (Source: [53])
Encounter Bubbles is a location aware social vizualisation software, which was presented
by Zhanna Shamis and Sean Savage in [53]. Encounter Bubbles is built on top of a location
based social service called Mobster. Only a PC software mock-up is developed yet, an early
prototype to prove the concept. A mobile version application is possibly going to be released in
the future. The software takes a similar approach to Jabberwocky, but relies on a centralized
platform and works with Wi-Fi hotspots for locating its users. With having all data stored
centrally, information can be provided to the user that Jabberwocky is not aware of, such as
looking at other users profiles. The user interface of the Encounter Bubbles software shows a
bubble for every person that the user encountered with, and sorts them in a chronological order.
Moreover, it can be scrolled to a specific time, whereupon the vizualisation enlarges the selected
point in time and displays it magnifying glass like. Screenshots of the user interface are shown
in figure 4 (having inverted colors for printing).
3.7
Multiple social network access and instant messaging
Fring [54] is a mobile application that allows to manage several social networks from only one application. Users can see if friends are online and send
messages. In contrast to mobile web interfaces of these social networks, a native application like Fring needs less bandwidth and is more responsive. Only communication to
the API has to be transmitted, not the complete website including the user interface. Fring also
supports several instant messaging protocols and voice-over-ip communication with a Skype or
SIP account.
Xumii [55] is a mobile social network, highly connected with other networks
like Facebook and MySpace and several instant messaging protocols. It offers mobile applications for J2ME enabled phones, iPhone OS, Symbian and
Blackberry OS, and a web interface for mobile browsers.
3
16
CLASSIFICATION
Nimbuzz [56] is a mobile social network, providing mobile applications for
nearly every platform: J2ME, Android, Symbian, Palm OS, Blackberry OS,
Windows Mobile and iPhone OS. It integrates with Facebook, MySpace, Hyves
and some more social networks and allows communication as well as photo sharing over these
social networks. Just like Fring, Nimbuzz has grants access to several instant messaging accounts
and Skype telephony.
3.7.1
Location aware versions
Buddymob [57] is a application for managing many profiles on several social
networks and instant messaging services. It has location aware features like
locating friends, or sending the own location to all logged in social networks.
3.8
Location based mobile social networks
Brightkite [58] is a well-established location aware social network, that provides a mobile web interface and a J2ME implemented application for most
mobile devices and supports several smart-phones by offering native applications for Android, Blackberry OS and the iPhone OS. It offers privacy by asking the user with
whom he likes to share the information when he enters it, not only in some nested setting menus.
The service offers possibilities to keep its users up-to-date in manifold ways: Users can see what
their friends do or what is happening around their current location. They can look at posts they
are mentioned in or read new comments on their posts.
Myrimis [59] is a location based mobile social network, predominantly used
in Asia. It offers to create geotagged photo albums. Myrimis can be accessed
by a mobile web interface or a J2ME or iPhone application.
3.8.1
Friend finding services
Loopt [60] is a location sensitive mobile social network since 2006. It is
one of the first networks to integrate a friend finding service based on the
location data updated by mobile applications. Those applications are available
for J2ME enabled phones, Android, Blackberry OS and the iPhone OS. Loopt
can connect to Facebook and Twitter accounts. Registration requires a US
mobile phone number, which is checked for its validity with a code sent by
SMS.
Google Latitude [61] is a location sensitive mobile social network launched
by Google in February 2009. It is targeted to friend finding, and provides native
mobile applications for all smart phone platforms, except for the Palm OS.
3
CLASSIFICATION
17
Moximity [62] is a location sensitive mobile social network, integrating
friends from other networks such as Facebook and Twitter with users’ location data. It overlays information about local venues such as restaurants and
bars, so users can see at which establishments their friends are currently located. A Moximity
mobile application exists for the iPhone OS.
MyGeolog [63] is a location aware mobile social network, based on a J2ME
implemented mobile application. The application allows for seeing nearby
friends on a map view, updating the users location and publishing of geo-tagged
photos and videos. It integrates with Facebook and Twitter and can even update Fireeagle (a
location storage service) with the latest location data.
Plazes [64] is a location based mobile social network, integrating with Facebook, Twitter and is able to publish the current location on Fireeagle. A mobile
application, called Plazer, is only available for the iPhone OS yet. Plazes has
a mobile web interface, with which users can place themselves or create venues, annotated with
the respective category. It is also possible to check-in the user’s position by sending a SMS with
the current address to the service as well as getting notified about updates via SMS.
Pocketlife [65] is a location aware mobile social network, letting people share
not only their positions, interesting places and geotagged pictures, but also
tracks. A J2ME mobile application, and one for the Blackberry OS and the
iPhone OS exist.
The Grid [66] is a South African location based mobile social network. Besides a friend finding service and the ability of posting geo-tagged photos, so
called ‘blips’, The Grid sends status updates on demand as twitter posts together with the current location. Mobile applications are offered as J2ME version as well as for
Android and the iPhone OS.
Locle [67] is a location aware mobile social network, primarily providing just
a friend finding service. It connects with Facebook and soon with MySpace and
Bebo [67]. It is one of the very few networks that offer a mobile application for
every single platform: J2ME, Android, Blackberry OS, iPhone OS, Symbian, Windows Mobile
and Palm OS.
Locatik [68] is a map based mobile social network, including a friend finding
service and the possibility to add places with photo and description. Users can
check-in their location via the mobile web interface, by sending a SMS with the
current address or by using the mobile application, only available for Symbian yet.
3
18
CLASSIFICATION
Buddy Beacon [69] is a location aware mobile social network, with the
primary target to be a friend finder service. It is a purely mobile application
for the platforms: Blackberry OS, Palm OS, Android, iPhone OS and all J2ME
capable platforms. It offers a map view to see the location of all friends, or a list view where
it shows the distance in relation to one’s own location. Users can of course decide whether to
update the location automatically or not or set themselves invisible to others.
Cometogethr [70] is a location based social network, with focus on mobile
usage. It provides just basic features of a classical social network service, like
having a friend-list, joining groups and messaging. It adds the location aware
feature of seeing current locations of friends.
Mologogo [71] is a location aware mobile social network, offering basically
a friend finding service via a mobile application, supporting J2ME phones, the
Blackberry OS and Windows Mobile.
3.8.2
Locating
Sniff [72] is a mobile friend finding service with social network integration
by a Facebook connection. The exceptional thing about this service is, that it
does not use locations that have to be updated by an application, but uses the
phone provider to determine the position by triangulation of cell towers within reach. Therefore
it delivers relatively inaccurate positions, but works in real time and without any software needed
on the mobile phone that ought to be localized.
Nowhere [73] is a German location based mobile social network, that allows
users to locate their friends by using cellular phone network providers’ positioning methods (For example U-TDOA). Nowhere users have to contract a
subscription allowing them to locate friends 15 times for a few Euros per week. These operations are independent from cell phone manufacturers or models and works with all German
cellular network providers. Located friends do not have to have any software installed on their
devices. But in order to be locatable, users have to be registered at Nowhere, be applied for the
locating service, have to be a friend of the user that wants to locate him and agree with getting
located by that particular user. Nowhere is accessible by a mobile web interface.
3.8.3
Tracking
Buddyway [74] is a location aware mobile social network with a rich tracking
service. With Buddyway, users can search for their friends locations, and can
let the service log their own location. By telling the system to start a ‘trip’,
the application starts sending its current location to the service in short time intervals. For this
purpose only accurate GPS locating makes sense, because of the much finer temporal and spatial
resolution than other techniques. The system then allows for adding geo-tagged annotations to
the trip like ‘coffee stop’ and displays information like duration of stops, or shows the point of
having had the maximum speed. The recorded trips can be shared with friends, or just stored
3
19
CLASSIFICATION
privately. Saved trips are displayed on a map, and additionally showing a height profile and a
diagram that shows the speed over a time axis.
iPoki [75] is a location based mobile social network, focused on real time GPS
tracking and friend finding services. Applications exist for Windows Mobile,
Blackberry OS and Symbian.
3.8.4
Meeting new people
Meetmoi [76] is a location tracking mobile social network for singles, offering
to alert its user when people come near that match the users filters for gender
and age. The service is accessible by a mobile web interface or by mobile
applications for Android, Blackberry OS, Windows Mobile and Symbian.
iPling [77] is a location aware mobile social network, focused on meeting new
people. Users tell their interests by entering tags. If users that have matching
interests come near, iPling displays them on a map, and the users can start a
conversation. The service is in a public beta test and supports the iPhone OS
and a mobile web interface.
Buddycloud [78] is a location based mobile social network that offers conversations in so called ‘channels’. These channels are similar to chat rooms or
bulletin boards in which users organize themselves in topic or interest related
groups. The service allows to share favorite places and the own location with others. The service is described as offering ”new ways to discover relevant people, places, and channels nearby”,
and is ”an extension of your social life beyond your close friends”. In contrast to most social
networks this service aims to get users to meet new people that share the same interests, and
not only staying in contact to already familiar friends.
Mobiluck [79] is a location aware mobile social network, based only on a
mobile web interface. Mobiluck has no mobile applications and no possibility
to automatically update the users’ positions. The current whereabouts have to
be either searched in the list of last used locations and favorite venues, or have to be entered as
an address to be geocoded by the service. Mobiluck aims to be a service for finding new friends
rather than only being a friend finding system. It offers a search for nearby Mobiluck users, or
users that had a look at the own profile. It allows for discovering new places, restaurants and
more, and integrates with Twitter.
3.8.5
Interest awareness
Match2blue [80] is a location- and interest-based mobile social networking
service. Users search match2blue for users who have identical or similar interests to themselves. Businesses can send offers to users that are interested.
Android and the iPhone OS are supported platforms of the mobile application.
3
CLASSIFICATION
20
Wenear [81] is a working concept of a location aware mobile social network,
not online for the general public yet. Wenear users are able to create locationand interest-tagged messages. A message will be received by other Wenear users
if they come near and the users interests match those set in the message. Immobile announcers
like shopping malls or other businesses, can broadcast announcements to Wenear users who get
in their proximity [82].
3.8.6
Friend finding, location sensitive content
IRL Connect [83] is a location based mobile social network, highly integrated with location based services. Facebook or Twitter users as well as
groups can be viewed on the map, as long as they contain geo information. Via
several map layers, users can overlay geographical positioned content such as news, Wikipedia
articles, last.fm [84] music events, Panoramio [85] pictures, YouTube [86] videos and some more.
The service only offers a mobile website and no application for any mobile devices, nevertheless
users can check-in their position to IRL Connect by using Loki [87] or Fireeagle [88] services
and their mobile applications.
Gypsii [89] is a location based mobile social network. Besides the friend
finding service and geo-tagged photo sharing, Gypsii has so called ‘Geo-bots’,
that can be added to the friend-list. These bots prepare location relevant news
content, to show up on the map of interesting places.
Qiro [90] is a German location and interest aware mobile social network,
offering its users to create ‘buttons’ they can attach to themselves, to show their
interests. These buttons are arranged in categories such as ‘music’, ‘movies’,
‘sports’ and some more. A bar vizualisation shows at first sight how many
buttons a user has in the different categories, represented by differently color coded bar pieces.
Moreover Qiro offers advertisers to promote their businesses by setting a position and a radius.
All users within this region are shown the promotion. Qiro shows not only friends on its map
view, but also geo relevant content taken from Qype [91], a location aware venue recommending
and review service. Qiro can be used with a mobile application from any J2ME capable mobile
device.
Urbian [92] ‘is about your hyper-local social network, the people you meet
in real life and the places you like’ [92]. It is a location based social network,
and features a friend finding service, as well as a database of places, entered by
users, and polls. Furthermore so called ‘hotspot modules’ can be added, these are map layers
with filtering mechanisms. The business module, for instance, allows for seeing all people who
work in a selected specific industry are shown on the map, the love module for finding dates etc.
Urbian provides the exceptional feature of having a ‘history’ function that shows how often the
user met one specific friend and for how long, or shows how often the user went to a specific
place in the last 6 months. Urbian is one of the only services that provide these statistic or
analysis features of saved geo data, even though it has just very basic query options and no
vizualisation yet.
3
21
CLASSIFICATION
Groovr [93] is a location sensitive social network. It features incoming filtering of messages by letting through either all, only messages of friends, or
only messages of the ‘inner circle’ of friends. Groovr describes itself as a friend
finder service and furnishes information about local events, sharing multimedia and is capable
of live chatting with friends.
Centrl [94] is a location based social network, that has mobile applications
that work on Blackberry OS, Android and iPhone OS. It lets users discover
locations of their friends, nearby strangers and geo-tagged offers or coupons
that can be created by restaurants, shops, bars, and other businesses. It provides login via one
of several social network accounts (it uses APIs of Facebook, MySpace, Hi5, Ning, Orkut, Bebo
or Friendster), so that users do not have to register once again to even one more social network.
It provides a integration of several third-party map layers like Yelp [95] (A recommendation and
review site), geo-marked Wikipedia articles and some more. These map layers are then shown
on the map view, together with locations of Centrl users. Centrl in turn provides an API itself,
to offer an interface for people to write their own application for Centrl, or to include Centrl in
their own systems as a service.
Nulaz [96] is a Dutch location based mobile social network. It connects with
Facebook, Twitter and Hyves. The map view shows not only nearby friends,
but also venues, pictures and geo-tagged content from Wikipedia and others.
Venues can be categorized into the topics business, party, shopping, local info
or friends, to enable users to filter them on the map view.
Scooble [97] is a German location based mobile social network and a friend
finder service. Beyond that, it provides a map view with geo-tagged content
from Wikipedia and Scooble users and rich filter possibilities. Scooble can get
directions to a specific place, or allows to manage photo albums and link pictures to users. A
mobile application only exists for the iPhone OS yet.
3.8.7
City Guide
Buzzd [98] was launched in 2008 and is a location based mobile social network
offering a mobile web-page and applications for the iPhone OS and Blackberry
OS. Buzzd has partnerships with content providers which place geo-tagged
information such as news, events, interesting venues and more on Buzzd’s map. Buzzd thus
represents a friend finding as well as a city guiding service.
3.8.8
Nightlife
MeetNowLive [99] is a North American location based mobile social network, targeted at people of cities and their nightlife plans and acquaintances.
It started with the two metropolises New York and Los Angeles and was now
extended by many other major cities in the United States. Besides seeing where their friends
are, users can browse events in their cities and search for happy hours and other offers, which are
3
22
CLASSIFICATION
shown on a map according to time range and distance filters. MeetNowLive mobile applications
are published for Blackberry OS, iPhone OS and Palm OS yet, however the service can be also
accessed by a mobile web interface.
3.8.9
Trips
Bliin [100] is a mobile social network that allows for managing and sharing
several geo-relevant information. It lets check in the users current location as
well as ‘spots’ (locations of interest), ‘trips’ (vacation residences or routes) or
geo-tagged multimedia. The service can be accessed by a mobile web browser
over a mobile web interface, as well as by an application that automatically refreshes the current
location. This application currently runs on the platforms J2ME, Blackberry OS, Windows
Mobile and iPhone OS. A full web interface for desktop PCs is also available, though the service
was developed mainly for mobile usage.
Wayn [101] is a location aware social network, orientated towards traveling.
It allows its users to plan trips and tell about their journeys in the past, together
with visited sites and venues placed on a virtual map, photos and videos. The
service has a virtual concurrency called ‘credits’ that can be used to send virtual gifts to friends
or get a higher rank right up to getting a VIP status. Higher ranks allow users to spice up their
profile pages with custom style templates. Credits can even be bought for real money, or just
earned by writing reviews about venues. Wayn just offers a mobile web-page and no mobile
application yet.
Rummble [102] is a location based mobile social network, with integration
of Facebook and Twitter connections. It allows to add trips consisting of start
and end date and a location. The map view shows places, so called ‘rummbles’,
which consist of a picture, a description, a position, involved people, the overall rating, comments
and are tagged with one ore more tags to be filterable. Rummble can be used by a web interface,
or a mobile application available for Android and the iPhone OS yet.
3.8.10
Alerting when friends are nearby
Dodgeball [103] was a SMS powered pure mobile social network. Dodgeball
was developed in 2000 by New York University students. The service could
receive SMS messages with text containing the users current location/address
and sent back a list of near friends and friends of friends, or interesting places around. It has
been acquired by Google in 2005 [104], and went offline because of the launch of Google Latitude,
that replaced the service completely.
3.8.11
Awarding systems
Skobbler [105] is a German location based mobile social network with a lot
of geo-tagging and recommending services. Skobbler encourages its users to
enter information about venues, rate them and give comments. Users can earn
3
CLASSIFICATION
23
virtual ‘karma’ points the more they participate in the community. Moreover the system divides
the world map in about two times two kilometer large grid squares. If a user has written more
articles about places of interest than any other user, he becomes the ‘Local Hero’ in that grid
square. A map view shows these grid squares with the current local hero and his challenger, red
squares indicate that a stranger has occupied this region, yellow squares are held by friends and
green ones are held by the user itself. These motivation methods are targetted to keep Skobbler
interesting, comparable to a game, and the service thus gets a rich database of places of interest,
that people can browse through.
Foursquare [106] is a mobile social network whose founders were partly involved with Dodgeball. The service leaves the impression like being a further
development of Dodgeball. Users can still check in their location using SMS
messages. Foursquare uses, however, more user interfaces like Dodgeball: a web interface, optimized for mobile browsers, and applications for Blackberry OS and iPhone OS platforms.
Foursquare motivates users to check in their location regularly by offering them badges. Moreover the system awards users, that check in more often than anyone else on one venue, the title
‘Mayor’ of this place. This awarding system is comparable to the ‘Local Hero’ awarding system
built into Skobbler. These systems playfully push users to quickly fill the services databases
with information to make the sites more popular.
3.8.12
Storytelling
Whrrl [107] is a location based mobile social network that is specialized on
letting people create so called ‘stories’. These stories are presentations with
several slides that can be filled with pictures and text. With these stories users
can ‘share and remember their real-world stories as they happen’ [108]. The story creator decides
which users may collaborate, and sets the visibility to ‘everyone’, ‘friends’, ‘trusted friends’ or
‘private’. Furthermore the story is geo-tagged to a specific location. Whrrl combines a photo
album with geotags and the possibility to letting many people collaborate and sharing it to
Facebook and Twitter, all from the mobile phone. The service features mobile applications for
J2ME platforms and the iPhone OS as well as a SMS based story publishing. SMS messages
can be sent to a certain phone number (94775) with several commands for creating stories,
extending them, or setting the status etc. This is why Whrrl has such a unusual short name
with only 5 letters: WHRRL = 94775. Of course Whrrl offers all common features of location
based networks like finding friends’ locations.
3.8.13
Location and distance aware bulletins
Zintin [109] is a purely mobile social network, that has a built in painting
tool in its mobile application. Shot photos can be edited, or new pictures can
be scribbled from scratch and shared with friends. Zintin offers the exceptional
feature of providing bulletin boards attached citys. ‘Pretty much every city in the world now
has its own wall on zintin. A city’s bulletin board is visible within a radius which grows with
its population, so you will be able to see small towns which are very nearby, and larger cities
which are a little further afield’ [110]. Zintin is only available for the iPhone OS so far.
3
CLASSIFICATION
3.9
24
Mobile social networks with microblogging
Blummi [111] is a service that is built on top of the Twitter API. It extends
Twitters microblogging service by adding several social networking features.
Users are able to not only ‘follow’ people, but also to manage friends in friendlists, and write private messages. It offers location based functions like recommending places
to friends or checking for current whereabouts of friends. It can also send a notification if a
friend comes nearby. Blummi is, in contrast to the following five hybrid systems, not a social
network that offers microblogging as an additional option, but is a microblogging service that
offers classical social networking features as an add on.
3.9.1
Interest and location aware
Trackut [112] is a location based mobile social network, offering friend finding
and tracking services. Users can share points of interests in channels, organized
by different interests. Users can subscribe to channels and get a notification
if they come near a venue that is listed in the channel, or if a friend comes nearby. Events,
that are associated with a location and a specific date and time can be created and shared with
friends. Mobile applications are currently available written in J2ME, and for the Android and
Blackberry OS platforms.
Socialight [113] is a location based mobile social network, organized in topicspecific channels, where users can leave geo-tagged ‘sticky notes’. The service
is similar to Trackut 3.9.1, but acts passively. Instead of notifying users when
they come nearby a note like Trackut does, Socialight lets look around its users if they search
for something specific.
3.9.2
Location based geojournals
Belysio [114] is a location aware social network that provides a web-page
optimized for mobile devices as well as a J2ME powered application for mobile phones. It allows for searching for people around the current location.
The current location can either be set manually by the user, or sent automatically by the application on the mobile device at times. Users can also create so called ‘geojournals’ to post
geo-tagged photos and messages, and share stories like vacation trips with friends. The service
allows to ‘follow’ other users, to get notifications if new geojournals are published. This feature
has a microblogging character, and therefore is categorized under ‘Mobile social networks with
microblogging’.
3.9.3
Multimedia sharing
Radar [115] is a mobile social network, connecting with Facebook and specialized on photo sharing. Users can look at their friends’ photos and profiles
by using the mobile web interface, or with the mobile application on a Blackberry or iPhone. Photos can be also uploaded via MMS, sent to a unique email address. Radar
4
EMPIRICAL ANALYSIS
25
provides such a email address for every user for that purpose. People can ‘follow’ other users’
photo streams in a microblog fashion.
Rabble [116] is a mobile social network with microblogging and photo sharing
functionality, accessible by a mobile web interface or the J2ME based mobile
application.
4
Empirical Analysis
Besides the characterization and classification of platform- and service-concepts, a natural goal
when characterizing Mobile Social Networking (MSN) is statistically characterizing the usage
patterns in the MSN universe. Unfortunately, as of September 2009, it is a complicating factor
that a large share of the more innovative platform- and service-concepts in MSN characterized
in the previous part 3 do not have a user basis that is broad enough for meaningful quantitative
studies. Furthermore, data about usage patterns of their mobile versions are hard to acquire.
We therefore first also reviewed and investigated the dynamics in conventional social networking
platforms having a mobile version, in order to be able to infer insights about potential dynamics
of their mobile versions, where necessary.
4.1
Social Network Crawler
For our investigations we selected a number of MSN platforms on the basis of a set of criteria:
• platforms that could also be accessed via non-mobile network in order to be able to efficiently investigate them.
• platforms that allowed for crawling the social network at all
• platforms that had a non trivial user-base and were ”alive´´ (had sufficient traffic (quantity
and up-to-date-ness)
• Facebook should be included as a reference
A crawler was written in C# that crawled the raw social network graph from several SNplatforms using anonymization for the profile- nodes. The application allows for comparative
visualization of the resulting graphs. For SN-platforms with numerical user-ids, random ids
were created and checked for validity. If valid, the corresponding network edges were crawled
from the respective profiles. For networks with string-based ids, we started from several known
user-ids and crawled the graph depth first until a depth of three. After that, duplicates were
removed. For both crawling styles we subsequently removed nodes with no adjacent edges.
Both approaches do not guarantee a strict statistical representativeness, but due to the small
world-like structure of social networks (see e.g. [117]), a sufficient degree of representativeness
to monitor the dynamics of these social networks can be assumed.
We crawled the networks from a selection of mobile social network platforms weekly for exactly
eight weeks, between 6. July and 31. August 2009, in order to analyze the dynamics of these
networks.
4
26
EMPIRICAL ANALYSIS
4.2
Network dynamics
Figures 5 and 6 shows the absolute and relative dynamics of the average number of friends
in several (mobile) social networks. The analysis reveals that on average some friends were
deleted, but in all examined systems users added more friends than they deleted in average.
The results also show, that sizes of friendlists highly depend on the size of the social network
in total. (Compare figure 7). It turns out that average friendlist sizes (average numbers of
Average number of friends
100,0
14,0
90,0
12,0
70,0
60,0
Facebook
Brightkite
50,0
Gypsii
40,0
Whrrl
Skobbler
30,0
Average number of friends
Average number of friends
80,0
Skobbler
10,0
Mobiluck
MeetNowLive
8,0
Plazes
Scooble
6,0
Trackut
Rummble
4,0
Bliin
20,0
Qiro
2,0
10,0
0,0
1
06.07.2009
0,0
2
3
03.08.2009
4
5
31.08.2009
1
06.07.2009
2
3
03.08.2009
4
5
31.08.2009
Figure 5: Dynamics of the average number of adjacent edges of a node in several mobile social
networks. (‘skobbler’ is plotted in both for orientation.)
Relative changes in friendlists
60,0%
50,0%
40,0%
30,0%
20,0%
friends added
friends deleted
10,0%
0,0%
Figure 6: Relative dynamics of average number of adjacent edges.
4
27
EMPIRICAL ANALYSIS
Numbers of registered users
Facebook
MySpace
Friendster
Hi5
Orkut
Twitter
Wayn
Hyves
Latitude
Mobiluck
Centrl
Nowhere
Whrrl
Skobbler
Trackut
Bliin
Qiro
Next2Friends
Scooble
Urbian
250.000.000
115.000.000
105.000.000
70.000.000
50.000.000
25.000.000
15.000.000
9.100.000
2.100.000
1.200.000
450.000
100000
40.000
25.000
23.000
22.600
12.000
11.000
1000
300
1E+10
10
billion
1E+09
1 billion
100000000
10000000
1000000
100000
10000
1000
100
10
1
Figure 7: Numbers of their registered users (August 2009), shown on a logarithmic scale.
adjacent nodes) scale relative to the absolute sizes of the overall social network of the respective
platforms. The obvious reason for this phenomenon is that for an average person the total share
of friends registered in a small platform is also small. This is one reason why the market for
mobile social networking is dominated by the existing large SN-platforms. Small and new mobile
social networks predominantly denote small growth, while popular networks grow even futher
(see figure 5)
Other reasons for startup difficulties of new innovative mobile SN platforms may be, that only
13% of all sold mobile phones in the first quarter of 2009 were smart phones [118] that provided
a basic UI (screen size, maturity of input techniques etc.) sufficient for both mobile web-browser
based or application based mobile SN services. In 2007 only 3.5% (USA) and 2.2% (Europe) of
all mobile device users used mobile social networking services [119]. Furthermore, sufficiently
cost effective flat-rates for mobile data-transfer are only becoming to be widespread as of August
2009, so that MSN is still quite costly, especially for the SMS-bases services.
Figure 8 shows that many users in small networks have nearly no friends connected in their
friendlists. Users either have left the service already or will do so, if the network does not offer
a genuine added value, compared to established large social networks. Many small networks are
built by early stage start-ups, without sufficient funds to start advertising and calling attention
to their services. Many of these are compelled to shut down their service after a while, and
many were shut down already.
Figure 8 complements the overview of figure 7 and further illustrates the findings just discussed,
4
28
EMPIRICAL ANALYSIS
100,000%
probability
10,000%
1,000%
Facebook
Brightkite
Whrrl
0,100%
0,010%
1
10
100
1000
number of friends
100,000%
GyPSii
Plazes
10,000%
Mobiluck
probability
Trackut
Rummble
1,000%
Scooble
Skobbler
MeetNowLive
0,100%
Qiro
Bliin
0,010%
1
10
100
1000
number of friends
Figure 8: The distribution of friend-list sizes of several classic social networks (with mobile
version) and distinctly mobile social networks. One data point denotes the probability of a user
having this specific number of edges (friends). Both diagrams are identically logarithmically
scaled.
showing the distribution of friend-list sizes. We see that small new mobile networks have distributions with a high probability of having very few friends which directly competing against the
established players gives them a small chance of sustained presence in the market.
4
29
EMPIRICAL ANALYSIS
4.2.1
Development of mobile social network start-ups
Since the end of the 1990s mobile social networking concepts enjoyed mixed success. Early MSN
platforms like the SixDegrees platform (launched 1997, ceased 2001) or Dodgeball (launched
in 2000) showed that while the market was not yet ready for these kind of concepts, some of
their ideas prevailed and were integrated in later platforms. The number of newly launched
mobile networks virtually exploded in 2007. Of all MSN platforms, the figure contains our
Social network launches
percentage of launches
35%
30%
25%
20%
15%
10%
5%
0%
1998
2000
2002
2004
2006
2008
Figure 9: Number of social networks launched in respective years. Fractions correspond to total
number of launches of social networks, discussed in this contribution.
complete selection of pure MSN platforms and only the most popular ”classic´´ SN platforms
with additional mobile interfaces. nearly all mobile interfaces and application for classic social
networks were developed after 2006. The launching dates were captured from the official blogs
or press releases of the network service developers or by searching the Internet Archive [120]
service. For this reason the dates may vary and may have only an accuracy of a whole year in
the worst case. The year 2009 was omitted because the year was not over until the completion
of this contribution.
4.3
Geographic dependencies in utilisation
The user-base of most platforms are concentrated in a certain region of the world. Region of
origin and region of highest polularity can differ: Orkut was developed in California, USA and
is now mainly used by Brazilian people [33].
4.4
4.4.1
Spread and usage of different mobile platforms
Usage of mobile platforms
In view of the question what OS platforms are used for mobile social networking, no direct usage
data are available for this contribution. Indirect reasoning may however may be accomplished
by analyzing the distribution of mobile internet traffic with respect to the various platforms (see
4
30
EMPIRICAL ANALYSIS
Asia
North America
Mixi, Multiply,
Orkut (India),
Wretch, Friendster,
Xiaonei, Cyworld
LinkedIn, MySpace, Twitter,
Nexopia (Canada)
South America
Orkut, Hi5
Europe
Bebo (UK), Hyves (NL), Hi5,
StudiVZ (Germany), MySpace, Twitter
Australia
Friendster, Bebo
Figure 10: Geographic dependencies of several social networks (data: [121] and own research)
Requests from smartphone OSs by origin
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
Other
Palm
Windows Mobile
Blackberry OS
Android
Symbian
iPhone OS
Figure 11: Percental fractions of requests from smartphone operating systems, broken down into
origin countries (numbers: AdMob Mobile Metrics Report 2009 [122]
figure 11). Sales figures of these platforms are in general not a good indicator, since the actual
internet usage may strongly depend on the tariffs chosen by the users.
Figure 11 illustrates this fact, by showing the six most popular platforms, all beeing smart phone
operating systems, and their proportion of total requests in the mobile internet of the respective
region. In Europe, for instance, iPhone users produce about 64% of all mobile internet requests
while the iPhone has just a market share of 1.5% of all cell phones. These are just 4 million
sold devices of total 269 million phones in the first quarter of 2009 [118]. Therefore, web request
numbers are much more significant than sales volumes.
4
31
EMPIRICAL ANALYSIS
The platforms Blackberry, Palm and Windows Mobile are very likely to find in business environments, whereas Symbian, the iPhone OS and Android in contrary are widely used for
entertainment. This may explain the fact that the three last mentioned platforms alone generate 85% of all mobile web requests, while business tasks only demand email and not web access.
Nearly all social networks target at entertainment and friend related usage patterns. This may,
in turn, be the reason for social network providers to not primarily develop mobile applications
for Palm and Windows Mobile operating systems. That, indeed, only few social networks support these systems, can be seen in figure 12.
RIMs Blackberry system seems to be the sole exception and is not only used for business purposes
but also a popular device for entertainment and social networking. 39% of all mobile social
networks provide an application for Blackberry, especially these which are domiciled in the US.
Having said that, Blackberry users in the US interestingly only make one percent of all internet
requests.
Supported operating systems
9%
4%
iPhone OS
23%
11%
J2ME capable OS
Blackberry OS
30%
39%
51%
51%
Android
Symbian
13%
23%
17%
25%
20%
9%
Windows Mobile
Palm OS
Figure 12: Distribution of available social network applications across the platforms. The small
diagrams indicate how many social networks offer an application for the particular operating
system.
Symbian and the iPhone operating system are apparently most popular worldwide for browsing
the internet, as figure 11 shows. Users of these two platforms also have a great variety of mobile
social software (see figure 12), considering that Symbian devices widely support J2ME applications.
Android, beeing the youngest of all six smartphone operating systems, is not this established
yet. Many devices providing this platform are still to come, but internet requests from Android
devices already amount to third most in the USA and Europe. 30% of all mobile social networks, presented in this thesis, already support the Android platform, as shown by the small
diagram, corresponding to Android, in figure 12. Moreover, several press releases or blogs of
social network providers state to deliver an Android based application for their services soon,
as for instance Gypsii [123].
Because every mobile social networking system can support different numbers of platforms, figure
5
32
CONCLUSION
13 points out how this is distributed and how many services offer total platform independence.
To provide a detailed overview over all presented mobile social networks and their supported
Number of existing mobile
applications per network
No application existing
13%
19%
14%
1 application for a single OS
2 operating systems
3 operating systems
12%
30%
4 operating systems
12%
5 operating systems or more
Figure 13: Distribution of social network platforms as either platform independent, dependent
on one to five special systems, or not offering any mobile application at all.
platforms, a tabular listing is shown in figure 14.
5
Conclusion
In this contribution we reviewed current mobile social networking platforms and services in order
to give an overview of the practical state of the art in this field. We discussed a classification
system for current MSN services and provided a short discussion and classification of the currently existing approaches. We regarded those MSN platforms that fulfilled certain maturity
criteria (e.g. have a non-trivial user-base (as of August 2009)). We then presented an empirical
analysis on MSN platforms, e.g. analyzing the dynamics of the underlying social networks, the
distribution with respect to mobile OS platforms used, the geographical distribution etc. Having
discussed current MSN systems and the functionality they provide, it is interesting to regard
possibilities and prerequisites for future services in the field of MSN.
5.1
Future approaches in MSN
Weighted and directed relations Introducing further attributes to the social relations such
as weights or a direction is an interesting possibility. In desktop-based SN platforms, weights
and directions can be deduced from statistical properties of communication acts (frequency, reciprocity, temporal variations (e.g. burst-behavior) etc. see e.g. [4]). In mobile social networking,
social context such as co-presence (e.g. modeled by social situations [124][125]) allows for an
even richer ground for deducing weights of social relations. Weighted relations could be used in
a variety of ways: Awareness services could focus on relations with lower weights for example,
privacy settings could be formulated according to edge weights etc.
33
Aka Aki
Areyouhere
Bebo
Belysio
Bliin
Blummi
Brightkite
Buddy Beacon
Buddycloud
Buddymob
Buddyway
Cellmate
Centrl
Cometogethr
Dodgeball
Foursquare
Friendster
Fring
Groovr
Grou.ps
GyPSii
Hi5
Hyves
iFob
iPling
iPoki
IRL Connect
Latitude
Locatik
Locle
Loopt
Mamjam
Match2Blue
MeetMoi
MeetNowLive
id
iPhon
e OS
Black
berry
OS
Wind
ows M
obile
Symb
ian
Palm
OS
Mob
ile op
timiz
ed w
SMS
eb
Andr
o
J2ME
e
Nam
id
iPhon
e OS
Black
berry
OS
Wind
ows M
obile
Symb
ian
Palm
OS
Mob
ile op
timiz
ed w
SMS
e
Andr
o
J2ME
Nam
e
binte
inter
face
CONCLUSION
rface
5
Mobiluck
Mologogo
Moximity
Mygeolog
Myrimis
MySpace
Next2Friends
Nimbuzz
Nowhere
Nulaz
Orkut
Plazes
Pocket Life
Qeep
Qiro
Rabble
Radar
Rummble
Scooble
Seamee
Shizzow
Skobbler
Sniff
Socialight
The Grid
Toai
Trackut
Urbian
Wayn
Wenear
Xumii
Zannel
Twitter
Buzzd
Figure 14: Listing of all discussed social networking services and their supported platforms and
interfaces.
Typified relations Typified social relations also have an interesting potential for new applications in SN and MSN. They may be used for privacy settings (e.g. restricting access to certain
parts of the personal information space only to users to whom a person has a relation of a
certain super-type), customizing awareness services or communication services (e.g. addressing
people with certain relations to the sender). Assigning social relations to one or more concepts
or types can either be done manually (as in a restricted way already realized by e.g. Facebook
groups) or potentially by analyzing the contents of communication between two parties. The
conceptualization itself can also be manually created (either by a platform provider or by a
certain group of users) or automatically deduced (e.g. as a folksonomy of relations from tags).
It can be a formal ontology (see e.g. [126][127]) or a simple flat or hierarchical taxonomy.
5
CONCLUSION
34
Individual and social context Mobile social networking allows for richer possibilities for
acquiring data on individual and social context. Examples have been discussed in section 1.
A typical example for individual context is location. Classes of services in individual mobile
information management (see e.g. [128]) that may use location include reminder-services, where
if, for instance, the user has an appointment, the system could check the user’s location a quarter
of an hour before. If the system then recognizes that the user is too far away for arriving in
time, it could send an automated excuse message to the other participants of the appointment.
A similar service already exists [129], and even additionally sends the estimated time of arrival.
Classes of services in MSN may especially use co-location in form of social situation models for
awareness- and communication services like Social Life Logging or SocioCast (see e.g. [125].
Especially interesting is the automatic deduction of social context on the basis of social signals
as discussed in 1 (see e.g. [124]). Location may be used in MSN for collaborative advertising and
information markets. The user could choose, that he wants to get the latest offers and happy
hours, if he is within some defined radius around bars and clubs. This context could be further
restricted by setting to receive offers by special users only or only at weekends or only within
a preferred time span at night. That would be a service delivering useful information, which is
passively received by the user, instead of having to search it every time. A similar technique,
for finding offers by specifying time ranges, is already used in the MeetNowLive mobile social
network, presented in section 3.8.8.
Advanced visualisations Orthogonal to the aforementioned possibilities, advanced visualization of and using context information is also an interesting aspect of future MSN services.
With most of today’s mobile social network applications only providing list view graphical interfaces, or at most a map view, more advanced user interfaces could appear in the future. Mobile
devices have very limited capabilities for output, due to the relatively small display sizes. Therefore it is a very challenging task to develop a user interface that shows as much information as
possible but simultaneously provides an overview and does not overstrain users. Approaches,
which ought to be named within this scope, are:
• 3D map views, letting the user orientate by nearby buildings, pictured on the map view.
• More advanced views of the current social context: Friend-finder services, dynamics of
parts of the personal social network (e.g. dynamics of the instantiations of certain social
relations in social situations [125]) etc.
• Timeline based visualisations (e.g. of parts of the social network), like already implemented
in Encounter Bubbles (see section 3.6.2). (see also e.g. [130]).
• Augmented reality views, already available for Android, within the mobile application Layar1 . This view avails itself of the camera, built into most mobile devices. With the devices
positioning abilities, its digital compass and accelerometer, it is able to detect where the
user stands and additionally where the camera is pointing to. This way the application can
overlay the cameras picture with markers, showing geo-tagged social networking content,
such as positions of friends, places of interest and more.
1
http://layar.eu/
REFERENCES
5.2
35
Completeness of this contribution
As a conclusion, after broad researches, it can be assured that nearly all relevant mobile social networks as of September 2009 were covered with this thesis, that provide extra features
compared to other networks. Surely, many small mobile social networks exist, which were not
presented. However, the popularity of such a system is often regionally restricted. Moreover
these systems, as far as they were discovered during the research, do not provide any added
value or additional functionality compared to respective systems, discussed in this thesis.
References
[1] Facebook. http://www.facebook.com/. (URL, Nov 2009).
[2] Myspace. http://www.myspace.com/. (URL, Nov 2009).
[3] Alexa top 500 global sites. www.alexa.com/topsites. (URL, Nov 2009).
[4] Georg Groh. Ad-Hoc-Groups in Mobile Communities - Detection, Modeling and Applications. PhD-Thesis, TU-Muenchen, 2005.
[5] Alessandro Vinciarelli, Maja Pantic, and Hervé Bourlard. Social signal processing: Survey
of an emerging domain. Image and Vision Computing, 27(12):1743 – 1759, 2009.
[6] A. Pentland N. Eagle and D. Lazer. Inferring social network structure using mobile phone
data. Proceedings of the National Academy of Sciences (PNAS) Vol 106(36), pp. 1527415278, 2009.
[7] A. Pentland N. Eagle. Eigenbehaviors: Identifying structure in routine. Behavioral Ecology
and Sociobiology 63:7, 2009.
[8] Mimi Sheller. Mobile publics: beyond the network perspective. Environment and Planning
D: Society and Space, 4:39–52, February 2004.
[9] Pedro G. Lind Marta C. González and Hans J. Herrmann. System of mobile agents to
model social networks. The American Physical Society, March 2006.
[10] Calabrese Francesco, Kloeckl Kristian, Ratti Carlo, Bilandzic Mark, Foth Marcus, Button
Angela, Klaebe Helen, Forlano Laura, White Sean, Morozov Petia, Feiner Steven, Girardin Fabien, Blat Josep, Nova Nicolas, Pieniazek M.P., Tieben Rob, van Boerdonk Koen,
Klooster Sietske, van den Hoven Elise, Serrano J. Martin, Serrat Joan, Michelis Daniel,
and Kabisch Eric. Urban computing and mobile devices. Pervasive Computing, IEEE,
6:52–57, July 2007.
[11] K. Fischbach D. Schoder and C. Schmitt. Core concepts in peer-to-peer networking: : The
evolution of a disruptive technology, 2005.
[12] Yamir Moreno Fang Wang and Yaoru Sun. Structure of peer-to-peer social networks.
Physical Review E, 73:036123, 2006., 2006.
[13] Noserub. http://noserub.com. (URL, Nov 2009).
REFERENCES
36
[14] Patrick Mukherjee Aleksandra Kovacevic Ralf Steinmetz Kalman Graffi, Sergey Podrajanski. A distributed platform for multimedia communities. IEEE International Symposium
on Multimedia (ISM ’08), p. 6, December 2008.
[15] Verena Rappel Georg Groh. Towards demarcation and modeling of small sub-communities
/ groups in p2p social networks. Proc. IEEE SocialCom09, Vancouver, Canada., 2009.
[16] Bonnie A. Nardi, Diane J. Schiano, Michelle Gumbrecht, and Luke Swartz. Why we blog.
Commun. ACM, 47(12):41–46, 2004.
[17] M. Baldauf, S. Dustdar, and F. Rosenberg. A survey on context-aware systems. International Journal of Ad Hoc and Ubiquitous Computing, 2(4):263–277, 2007.
[18] Twitter. http://www.twitter.com/. (URL, Nov 2009).
[19] John Breslin and Stefan Decker. The future of social networks on the internet: The need
for semantics. IEEE Internet Computing, 11:86–90, 2007.
[20] John Breslin, Alexandre Passant, and Stefan Decker. The Social Semantic Web. SpringerVerlag, Heidelberg, 2009.
[21] Web2null. http://www.web2null.de/. (URL, Nov 2009).
[22] Gomo news. http://www.gomonews.com/moso/. (URL, Nov 2009).
[23] Claudio Schapsis.
Location based social networking links. http://bdnooz.com/
lbsn-location-based-social-networking-links/, 2009. (URL, Nov 2009).
[24] M-Trends. Mososo and wi-fi. http://www.m-trends.org/2005/11/mososo-wi-fi.html,
November 2005. (URL, Nov 2009).
[25] Elasticspace.
Mobile social software.
http://www.elasticspace.com/2004/06/
mobile-social-software, June 2004. (URL, Nov 2009).
[26] Friendster. http://www.friendster.com/. (URL, Nov 2009).
[27] Friendster. Press release. http://www.friendster.com/info/presscenter.php?A=
pr36, February 2008. (URL, Nov 2009).
[28] Opensocial. http://www.opensocial.org. (URL, Nov 2009).
[29] Hi5. http://hi5.com/. (URL, Nov 2009).
[30] USA Today. Social-networking sites going global. http://www.usatoday.com/money/
industries/technology/2008-02-10-social-networking-global_N.htm,
February
2008. (URL, Nov 2009).
[31] Hyves. http://www.hyves.nl/. (URL, Nov 2009).
[32] Orkut. http://www.orkut.com/. (URL, Nov 2009).
[33] Orkut. Demography statistics page. http://www.orkut.com/Main#MembersAll, September 2009. (URL, Nov 2009).
REFERENCES
37
[34] Bebo. http://www.bebo.com/. (URL, Nov 2009).
[35] BBC News. Aol acquires bebo social network. http://news.bbc.co.uk/1/hi/business/
7294174.stm, March 2008. (URL, Nov 2009).
[36] m-bebo. http://m.bebo.com. (URL, Nov 2009).
[37] Alexa. http://www.alexa.com/siteinfo/twitter.com. (URL, Nov 2009).
[38] Zannel. http://www.zannel.com/. (URL, Nov 2009).
[39] Shizzow. http://www.shizzow.com/. (URL, Nov 2009).
[40] Moko mobi. http://www.moko.mobi/. (URL, Nov 2009).
[41] Qeep. http://www.qeep.net/. (URL, Nov 2009).
[42] Aka-aki. http://www.aka-aki.com/. (URL, Nov 2009).
[43] Seamee. http://www.seamee.com/. (URL, Nov 2009).
[44] Next2friends. http://next2friends.com/. (URL, Nov 2009).
[45] Gomo news. http://www.mcliques.com/product.html. (URL, Nov 2009).
[46] Are you here? http://www.areyouhere.net/. (URL, Nov 2009).
[47] Clicmobile. http://www.clicmobile.fr. (URL, Nov 2009).
[48] Bluetooth. http://www.bluetooth.com. (URL, Nov 2009).
[49] ifob. http://www.icloseby.com/what_is_ifob.html. (URL, Nov 2009).
[50] Jabberwocky - about page. http://www.urban-atmospheres.net/Jabberwocky/about.
htm. (URL, Nov 2009).
[51] Jabberwocky - demo page. http://www.urban-atmospheres.net/Jabberwocky/demo.
htm. (URL, Nov 2009).
[52] Jabberwocky - information page. http://www.urban-atmospheres.net/Jabberwocky/
info.htm. (URL, Nov 2009).
[53] Encounter bubbles. http://www.seansavage.com/encounter-bubbles/. (URL, Nov
2009).
[54] Fring. http://www.fring.com/. (URL, Nov 2009).
[55] Xumii. http://www.xumii.com/. (URL, Nov 2009).
[56] Nimbuzz. http://www.nimbuzz.com/. (URL, Nov 2009).
[57] Buddymob. http://www.buddymob.com/. (URL, Nov 2009).
[58] Brightkite. http://brightkite.com/. (URL, Nov 2009).
[59] Myrimis. http://www.myrimis.com/. (URL, Nov 2009).
REFERENCES
[60] Loopt. http://www.loopt.com/. (URL, Nov 2009).
[61] Latitude. http://www.google.com/latitude/intro.html. (URL, Nov 2009).
[62] Moximity. http://www.moximity.com/. (URL, Nov 2009).
[63] Mygeolog. http://www.mygeolog.com/. (URL, Nov 2009).
[64] Plazes. http://plazes.com/. (URL, Nov 2009).
[65] Pocketlife. http://pocketlife.com/. (URL, Nov 2009).
[66] The grid. http://www.thegrid.co.za/. (URL, Nov 2009).
[67] Locle. http://www.locle.com/. (URL, Nov 2009).
[68] Locatik. http://www.locatik.com/. (URL, Nov 2009).
[69] Buddybeacon. http://where.com/buddybeacon/. (URL, Nov 2009).
[70] Cometogethr. http://www.cometogethr.com/. (URL, Nov 2009).
[71] Mologogo. http://www.mologogo.com/. (URL, Nov 2009).
[72] Sniff. http://www.sniffu.com/us/. (URL, Nov 2009).
[73] Nowhere. http://www.nowhere.de/. (URL, Nov 2009).
[74] Buddyway. http://www.buddyway.com/. (URL, Nov 2009).
[75] ipoki. http://www.ipoki.com/. (URL, Nov 2009).
[76] Meetmoi. http://www.meetmoi.com/. (URL, Nov 2009).
[77] ipling. http://www.ipling.com/. (URL, Nov 2009).
[78] Buddycloud. http://buddycloud.com/. (URL, Nov 2009).
[79] Mobiluck. http://www.mobiluck.com/. (URL, Nov 2009).
[80] Match2blue. http://match2blue.com/. (URL, Nov 2009).
[81] Wenear. http://www.wenear.com/. (URL, Nov 2009).
[82] Wenear. How? http://wenear.com/how, August 2009. (URL, Nov 2009).
[83] Irl connect. http://irlconnect.com/. (URL, Nov 2009).
[84] Lastfm. http://www.last.fm. (URL, Nov 2009).
[85] Panoramio. http://www.panoramio.com. (URL, Nov 2009).
[86] Youtube. http://www.youtube.com. (URL, Nov 2009).
[87] Loki. http://www.loki.com. (URL, Nov 2009).
[88] Fireeagle. http://fireeagle.yahoo.net. (URL, Nov 2009).
38
REFERENCES
39
[89] Gypsii. http://www.gypsii.com/. (URL, Nov 2009).
[90] Qiro. http://www.qiro.net/. (URL, Nov 2009).
[91] Qype. http://www.qype.com. (URL, Nov 2009).
[92] Urbian. http://www.urbian.org/. (URL, Nov 2009).
[93] Groovr. http://www.groovr.com/. (URL, Nov 2009).
[94] Centrl. http://centrl.com/. (URL, Nov 2009).
[95] Yelp. http://www.yelp.com. (URL, Nov 2009).
[96] Nulaz. http://www.nulaz.net/. (URL, Nov 2009).
[97] Scooble. http://scooble.de. (URL, Nov 2009).
[98] Buzzd. http://buzzd.com/. (URL, Nov 2009).
[99] Meetnowlive. http://www.meetnowlive.com/. (URL, Nov 2009).
[100] Bliin. http://www.bliin.com/. (URL, Nov 2009).
[101] Wayn. http://www.wayn.com/. (URL, Nov 2009).
[102] Rummble. http://www.rummble.com/. (URL, Nov 2009).
[103] Dodgeball. http://www.dodgeball.com/. (URL, Nov 2009).
[104] Nina D. Ziv and Bala Mulloth. An exploration on mobile social networking: Dodgeball as
a case in point. Proc. IEEE MBusiness Conference, June 2006.
[105] Skobbler. http://beta.skobbler.de/. (URL, Nov 2009).
[106] Foursquare. http://playfoursquare.com/. (URL, Nov 2009).
[107] Whrrl. http://whrrl.com/. (URL, Nov 2009).
[108] Whrrl. Faq. http://faq.whrrl.com/, August 2009. (URL, Nov 2009).
[109] Zintin. http://www.zintin.com/. (URL, Nov 2009).
[110] Zintin.
Bulletin board feature.
http://www.zintin.com/blog/2008/07/
new-feature-city-bulletin-boards/, August 2009. (URL, Nov 2009).
[111] Blummi. http://www.blummi.com/. (URL, Nov 2009).
[112] Trackut. http://www.trackut.com/. (URL, Nov 2009).
[113] Socialight. http://socialight.com/. (URL, Nov 2009).
[114] Belysio. http://www.belysio.com/. (URL, Nov 2009).
[115] Radar. http://radar.net/. (URL, Nov 2009).
[116] Rabble. http://www.rabble.com/. (URL, Nov 2009).
REFERENCES
40
[117] Duncan J. Watts and Steven H. Strogatz. Collective dynamics of /‘small-world/’ networks.
Nature, 393(6684):440–442, 1998.
[118] Gartner Inc. Press release. http://www.gartner.com/it/page.jsp?id=985912, May
2009. (URL, Nov 2009).
[119] M:Metrics. Mobile social networking. http://www.mmetrics.com/press/articles/
20070815-socialnetworking.pdf, August 2007. (URL, Nov 2009).
[120] Archhive. http://www.archive.org. (URL, Nov 2009).
[121] Wikipedia. Social network service. http://en.wikipedia.org/wiki/Social_Network_
Service, September 2009. (URL, Nov 2009).
[122] Admob. Mobile metrics. http://metrics.admob.com/wp-content/uploads/2009/08/
AdMob-Mobile-Metrics-July-09.pdf, July 2009. (URL, Nov 2009).
[123] Network World. Gypsii plans to bring android version. http://www.networkworld.com/
news/2009/090809-gypsii-offers-java-version-of.html, August 2009. (URL, Nov
2009).
[124] Georg Groh, Alexander Lehmann, Jonas Reimers, Rene Friess, and Loren Schwarz. Detecting social situations from interaction geometry. Proc IEEE SocialCom2010, Minneapolis
USA, August 2010.
[125] Georg Groh, Alexander Lehmann, Tianyu Wang, Stefan Huber, and Felix Hammerl. Applications for social situation models. Proc IADIS WAC 2010, Freiburg, Germany, July
2010.
[126] Peter Mika and Aldo Gangemi. Descriptions of social relations. Proceedings of the 1st
Workshop on Friend of a Friend, Social Networking and the (Semantic) Web, 2004, 20004.
[127] Foaf. http://www.foaf-project.org. (URL, Nov 2009).
[128] Wolfgang Woerndl, Georg Groh, and Aleksandar Hristov. Individual and social recommendations for mobile semantic personal information management. Journal on Advances
in Information Technology, 2(2), 2009.
[129] Sebastian Gros̈. Mobile innovation report. Pixelpark Whitepaper, 2009.
[130] Georg Groh, Holger Hanstein, and Wolfgang Woerndl. Interactively visualizing dynamic
social networks with dyson. Workshop Visual Interfaces to the Social and the Semantic
Web (VISSW 2009), IUI Conf., Sanibel Island, FL, Feb 2009.
[131] Anthony LaMarca Jeffrey Hightower and Ian E. Smith. Practical lessons from place lab.
Pervasive Computing, IEEE, 2006.
[132] Jörn Davidsen Holger Ebel and Stefan Bornholdt. Dynamics of social networks. Complexity, 8(2):24–27, December 2006.
[133] Sara Metcalf and Mark Paich. Spatial dynamics of social network evolution. 23rd International Conference of the System Dynamics Society, 2005.
REFERENCES
41
[134] Zuoliang Chen and Shigeyoshi Watanabe. A case study of applying sna to analyze cscl social network. Seventh IEEE International Conference on Advanced Learning Technologies,
2007.
[135] Elizabeth F. Churchill and Christine A. Halverson. Social networks and social networking.
IEEE Internet Computing, 2005.
[136] Thayne R. Coffman and Sherry E. Marcus. Pattern classification in social network analysis:
A case study. IEEEAC paper 1090, 2004.
[137] Tanya Y. Berger-Wolf and Jared Saia. A framework for analysis of dynamic social networks. In KDD ’06: Proceedings of the 12th ACM SIGKDD international conference on
Knowledge discovery and data mining, pages 523–528, New York, NY, USA, 2006. ACM.
[138] Chang Tong. Analysis of some popular mobile social network systems. Seminar on Internetworking, Helsinki University of Technology, 2008.
[139] Scott Counts and Karen E. Fisher. Mobile social networking: An information grounds
perspective. Proc. 41st Hawaii International Conference on System Sciences, 2008.
[140] Sebastian Ammermueller Oliver Bohl, Shakib Manouchehri and Oliver Gerstheimer. Mobile social software - potentials and limitations of enabling social networking on mobile
devices. Proc. Sixth International Conference on the Management of Mobile Business,
2007.
[141] Mehdi Mani, Anh-Minh Ngyuen, and Noel Crespi. What’s up: P2p spontaneous social
networking. Pervasive Computing and Communications, IEEE International Conference
on, 0:1–2, 2009.
[142] Alfred C. Weaver and Benjamin B. Morrison. Social networking. Computer, 41(2):97–100,
2008.
[143] Nathan Eagle and Alex Pentland. Social serendipity: Mobilizing social software. IEEE
Pervasive Computing, 4(2):28–34, 2005.
[144] Scott Counts and Marc Smith. Where were we: communities for sharing space-time trails.
In GIS ’07: Proceedings of the 15th annual ACM international symposium on Advances
in geographic information systems, pages 1–8, New York, NY, USA, 2007. ACM.